How to enable Azure Open AI?

About this article

This article will explain how to enable Azure Open AI connector and provide end-users access to the AI text assistant in the Library pane. The AI text assistant will help the users to summarize, rephrase, improve, or elaborate existing text in Word, PowerPoint, and Outlook. 

Prerequisites

How to set up Azure Open AI in Templafy?

1. In the Admin Center, access Integrations section
2. Go in Available tab
3. Find/Search for Azure Open AI
4. Click +
5. Enter the required information
   • Display Name
   • Words that describe the correct tone of voice
   • Words that describe off-brand language
6. Click Save

Security and Data privacy information

• The AI text assistant will process all information that is actioned by the AI text assistant. This means that if PII (Personal Identifiable Information) is part of the text, this information will be processed. No information fed into the tool will be used to train the AI learning engine.
• The prompts and completions data may be temporarily stored by the Microsoft Azure OpenAI Service in the European Economic Area for up to 30 days. This data is encrypted and is only accessible to authorized Microsoft employees for (1) debugging purposes in the event of a failure, and (2) investigating patterns of abuse and misuse to determine if the service is being used in a manner that violates the applicable product terms. Note: When a customer is approved for modified abuse monitoring, prompts and completions data are not stored, and thus Microsoft employees have no access to the data. For more information, click here.
• There are no third-parties involved in the processing besides Templafy and Microsoft Azure. No data is stored on the side of Templafy, all input and output data for this service is passed directly through Microsoft Open AI.
• Templafy will improve and enhance our Service by tracking usage data during the beta period as stated in the beta agreement. Please note user provided content will not be tracked or logged. The usage data is pseudonymized.

Related articles